Nvidia (NVDA) reported its fiscal first quarter results after the bell on Wednesday, topping analysts’ expectations on the top and bottom lines and offering a better-than-anticipated Q2 outlook.
Nvidia stock initially fell more than 2% on the news.
For the second quarter, Nvidia said it anticipates revenue between $89.1 billion and $92.8 billion. Wall Street was looking for $87.3 billion.
For Q1, Nvidia reported earnings per share (EPS) of $1.87 on revenue of $81.62 billion, beating the $1.77 and $79.18 analysts had expected. The company also boosted its quarterly dividend to $0.25 per share.
Nvidia’s data center business brought in the bulk of its revenue, topping out at $75.2 billion. Wall Street analysts were projecting $73.47 billion. That’s up from the $39.11 billion Nvidia reported in Q1 last year.
According to CFO Colette Kress, hyperscaler revenue accounted for 50% of data center sales in the quarter. The remaining 50% came from a variety of sources including “AI Clouds, industrial, enterprise, and sovereign customers.”
Nvidia did not see any revenue from Hopper products into China occurred during the quarter.
Nvidia is also changing the way it reports its earnings. The company is now breaking down its sales into Data Center and Edge Computing segments. Data center will include hyperscalers and AI clouds, industrial, and enterprise sources.
Edge Computing will be comprised of devices or agentic and physical AI, which the company says will include “PCs, game consoles, workstations, AI-RAN base stations, robotics, and automotive.”
Nvidia is contending with increasing competition in the AI processor space, including from Cerebras (CBRS), which held its initial public offering last Thursday.
Cerebras sells a different form of AI processor than Nvidia, which the company says offers faster overall performance thanks to its unique design.
In its latest earnings report, Amazon announced that its chip segment now has an annual revenue run rate of more than $20 billion and is growing at triple-digit percentages year over year.
The company also said it signed a deal with OpenAI (OPAI.PVT) via AWS to provide 2 gigawatts of capacity for its Trainium chips and that Anthropic (ANTH.PVT) agreed to use up to 5 gigawatts of current and future Trainium chips.
And Tuesday, Google touted its new TPU 8i and TPU 8t chip as part of its annual Google I/O conference in Mountain View, Calif. The TPU 8i is intended for running AI software, while the TPU 8t is designed for training models.
Like Amazon, Google has also signed a multigeneration, multigigawatt deal to provide its TPU chips to Anthropic, The Information reported.
